Abstract
Recent advances in the statistical methodology for selecting optimal subsets of features for data representation and classification are presented. This chapter attempts to provide a guideline of which approach to choose with respect to the extent of a priori knowledge of the problem. Two basic approaches are reviewed and the conditions under which they should be used are specified. One approach involves the use of the computationally effective Floating search methods. The alternative approach trades off the requirement for a priori information for the requirement of sufficient data to represent the distributions involved. Owing to its nature it is particularly suitable for cases when the underlying probability distributions are not unimodal. The approach attempts to achieve simultaneous feature selection and decision rule inference. According to the criterion adopted there are two variants allowing the selection of features either for optimal representation or discrimination.
Supported by the grants of Czech Ministry of Education MŠMT No.VS96063, Czech Acad.Sci. A2075608 and Grant Agency of the Czech Republic No.402/97/1242.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Boekee, D. E. and Lubbe, J. C. A. V. D. (1979). Some aspects of error bounds in feature selection. Pattern Recognition, 11:353–360.
Devijver, P. A. and Kittler, J. (1982). Pattern Recognition: A Statistical Approach. Prentice-Hall.
F.J. Ferri, P. Pudil, M. Hatef, and J. Kittler (1994). Comparative study of technique for large-scale features selection. In Pattern Recognition in Practice IV: Multiple Paradigms,Comparative Studies and Hybrid Systems; edited by E.S. Gelsema and L.N.Kanal,pages 403–413, North-Holland Elsevier.
Jain, A. K. and Zongker, D. (1997). Feature selection: Evaluation, application and small sample performance. IEEE Transactions on PAMI,19:153–158.
Novovičová, J. and Pudil, P. (1997). Dealing With Complexity: Neural Network Approach, chapter Feature Selection and Classification by Modified Model with Latent Structure. Springer Verlag.
Novovičová, J., Pudil, P., and Kittler, J. (1996). Divergence based feature selection for multimodal class densities. IEEE Transactions on Pattern Analysis and Machine Intelligence, 18:218–223.
Pudil, P. and Novovičová, J. (1998). Novel methods for subset selection with respect to problem knowledge. IEEE Intelligent Systems - Special Issue on Feature Transformation and Subset Selection,in print.
Pudil, P., Novovičová, J., Choakjarernwanit, N., and Kittler, J. (1993). An analysis of the max-min approach to feature selection. Pattern Recognition Letters, 14(11):841–847.
Pudil, P. Novovičová, J. Choakjarernwanit, N., and Kittler, J. (1995). Feature selection based on the approximation of class densities by finite mixtures of special type. Pattern Recognition,28(9):1389–98.
Pudil, P., Novovičová, J., and Kittler, J. (1994). Floating search methods in feature selection. Pattern Recognition Letters, 15:1119–1125.
Redner, R. A. and Walker, H. F. (1984). Mixture densities, maximum likelihood and the EM algorithm. SIAM J. Appl. Math.,26(2):195–239.
Siedlecki, W. and Sklansky, J. (1988). On automatic feature selection. International Journal of Pattern Recognition and Artificial Intelligence, 2(2):197–220.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1998 Springer Science+Business Media New York
About this chapter
Cite this chapter
Pudil, P., Novovičová, J. (1998). Novel Methods for Feature Subset Selection with Respect to Problem Knowledge. In: Liu, H., Motoda, H. (eds) Feature Extraction, Construction and Selection. The Springer International Series in Engineering and Computer Science, vol 453. Springer, Boston, MA. https://doi.org/10.1007/978-1-4615-5725-8_7
Download citation
DOI: https://doi.org/10.1007/978-1-4615-5725-8_7
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4613-7622-4
Online ISBN: 978-1-4615-5725-8
eBook Packages: Springer Book Archive